Executive Summary
Adversa AI's GuardFall research found the command safety check in ten of eleven open-source coding and computer-use agents could be bypassed using a long-known shell technique; only Continue was built to resist it.
What Happened
The safety check that is supposed to stop an AI coding agent from running a dangerous command can be walked straight past using a shell trick that has been public for decades. New research from Adversa AI, which is named the bypass GuardFall, found it works against ten of the eleven popular open-source coding and computer-use agents the firm tested. Only one, "Continue," was built to
Affected Products
- Open-source AI coding agents
Vulnerability / Attack Technique
Bypass of agent command safety checks
Technical Mechanism
As reported, the agents' command filters did not account for a decades-old class of shell behaviour, so blocked commands could still run.
AI Security Category
LLM06:2025 Excessive Agency
Part of the AI Agent Excessive Agency topic.
Relevant Frameworks
RealGround Analysis
Text-based command filters are fragile. RealGround analysis: isolate the agent's execution environment instead of relying on command matching.
Recommended Controls
- Run coding agents in containers with limited permissions
- Update agents to versions with fixed command handling
Related RealGround Service
Secure AI Agent Build · or start with the free AI Agent Security Scanner.
Sources
thehackernews.com: https://thehackernews.com/2026/06/guardfall-exposes-open-source-ai-coding.html
Summary and analysis written by RealGround from the source above; last updated 2026-10-18.